ANALYSIS OF PERSON ’S PROBLEM-SOLVING SKILLS

,


INTRODUCTION
21st-century learning places greater emphasis on project-based, problem-based, inquiry, design, and discovery (Fazrul Prasetya Nur Fahrozy, 2022).Students learn science to be able to create a product that is planned through existing problems, then carry out investigations to solve the solution.Then students will find a solution related to the problem as a product.The student's results will be able to have a positive influence on him to face future challenges and will provide solutions to the problems currently being faced.
Students' skills that are of concern in 21st-century learning are known as 4C skills (Yose Indarta, 2021), namely: critical thinking and problem-solving, creativity and discovery, communication and collaboration.Students will learn directly and think critically when they encounter problems, and then look for solutions.The solution is poured into a product that is useful and used for the future.
Improving critical thinking and problem-solving skills can be done through industrial job training (Shazaitul Azreen Rodzalan, 2020).Through industrial job training, students have several advantages.This requires students to have critical thinking skills to solve problems.
Students can reflect on the knowledge they have from college.
Problem-solving skills in mathematics learning have an important role, especially in learning that is linked to the real world, so students have challenges in solving these problems (Nurdan Özreçberoğlu, 2018).Problem-solving skills in mathematics learning prioritize nonroutine questions.This is because students become confident, independent, and think creatively.
The factor that influences students' lack of problem-solving skills is the teacher's way of teaching.Teachers still use teacher-centered learning methods.Although many methods can help students improve their problem-solving skills (Munali, 2023).
The results of measuring OSN (National Science Olympiad) critical thinking skills of class VIII students consisting of 24 students, namely: 9 male students and 15 female students show that: students' ability with interpretation sub-skills is 63% in the good category.Meanwhile, sub-skills: analysis, evaluation, inference, explanation, and self-regulation are less than 50% in the low category (Hasan Basri, 2019).Observation skills indicators: 1) collecting information, 2) understanding and interpreting meaning by identifying keywords, and 3) observing patterns, similarities, and differences in a problem.
Critical thinking skills indicators: 1) conceptualizing skills are the ability to identify, recognize, and investigate problems partially and completely; 2) logical reasoning is the ability to retrieve some relevant information, interpret the information, provide arguments, and provide conclusions; 3) application skills are the ability to use knowledge in new or familiar situations to solve problems through facts, knowledge, principles, and techniques; 4) analytical thinking is the ability to analyze arguments, prove, interpret, assess or evaluate, make conclusions, 5) decision-making skills are the ability to make choices, solve problems, make effective decisions based on reasons and arguments right; and 6) synthesis skills is the ability to combine parts to make them interrelated and then these parts become something new.Data processing was carried out using the Rasch Model.

METHODS
The Rasch Model is a data analysis technique developed to increase the accuracy of the instruments being developed, assess the quality of the instruments, and identify persons' abilities (Andrich, 1989;Davidowitz, 2016).By using raw data in the form of dichotomous or polytomous data, you can identify persons' abilities and the level of difficulty of the test items (Lutfi Nur, 2020).The Rasch Model formulates a model that connects people and items on the same logit scale (Rika Irmayanti, 2023).
The Rasch Model is also useful as a diagnostic tool to determine the reliability of instruments and the validity of test items (Hye Sun You, 2017 Testing the suitability of the Rasch Model data by examining items and persons (Shiau Wei Chan, 2020): the score for one of the following assumptions is met: 0.50 < Outfit MNSQ < 1.50; -2.00 < Outfit ZSTD < +2.00; or 0.40 < Point Measure Correlation < 0.85.
Data analysis using the Rasch Model can inform question items and people (Sumintono, 2013): 1) level of suitability of question items (item fit order), 2) level of difficulty of question items (item measure), 3) DIF detection of question items, 4) level person suitability (person fit order), 5) person ability level (person measure), 6) instrument analysis: (a) person measure, (b) Cronbach's Alpha, (c) person reliability and item reliability, and (8) test information function.
In this study, researchers analyzed students' abilities in responding to problem-solving skills questions: 1) Cronbach's Alpha person reliability, 2) person reliability, 3) person mean measure, and 4) measurement information function.

1) Fit Person
Before analyzing a person's abilities, the first step is to determine the person's fit against the Rasch Model.After analysis, there were 271 fit persons out of 298 persons.There are 27 misfit persons for the Rasch Model.The results of the misfit person analysis can be seen in Table 2 below:  , 80, 246, 260, 239, 110, 137, 129, 38, 128, 193, 82, 271, 18, 96, 37, 130, 86, 174, 8, 39, 272, 278, 298, 131, 6, 200 A total of 27 misfit persons were dropped from the data, then the data was analyzed again.After the data was reanalyzed, a total of 271 persons were fit for the Rasch Model.
Some of the fit person data can be seen in Table 3 below.The criteria used to check a person's fit with the Rasch Model are that one of the Outfit MNSQ, Outfit ZSTD, or Point Measure Correlation scores are met.If a person does not meet these three criteria, then the person is not good enough and needs to be repaired or replaced.
Based on Table 3, it can be concluded: 1) the person's ability is more than 0.00 logit, namely person number 109 is 0.23 logit and person number 106 is 0.06 logit; 2) the lowest person's ability is person number 7 is -3.807 logit and person number 14 is -4.21 logit.

2) Person Reliability
Person reliability can be seen in Table 4 below: The mean person measure score shows the mean score of all persons in responding to the question items.The mean person measure score of less than 0.00 logit indicates that the person's ability is less than the difficulty level of the item.Based on Table 4, the mean person measure score is -1.16 logit < 0.00 logit, indicating that the person's ability is less than the difficulty level of the questions.

3) Test Information Function
Each measurement will produce information related to the measurement results.The measurement information function describes the person's ability to respond to the item as a whole.The higher the peak of the test information function, the higher the measurement information function (Erina Afriani, 2023).The higher the measurement information function is related to the person's ability to respond to the question items.The information function of the problem-solving skills test can be seen in Figure 1 below.Figure 1 shows that the X axis is the person's ability to respond to problem-solving skills questions and the Y axis is the information function score.It can be seen that at low abilities, the information obtained from measurements is low.Likewise, for high abilities, the information obtained from measurements is also low.
The person's ability lies between -9.00 logit to 4 logit, which is -2.50 logit in the low category.The maximum information function is 22%, providing information that the correct accuracy of measuring problem-solving skills items is 22% in the low category.

CONCLUSION
The accuracy of the person's ability to respond to problem-solving skills questions was 22%.Based on these results, it can be concluded: that students' problem-solving skills are in the low category.Solutions that can be taken to improve students' problem-solving skills are getting used to responding to mathematical problems related to everyday life.In the next research, it is hoped that researchers will be able to measure students' problem-solving skills again using relevant courses.This measurement is expected to have student abilities of more than 22%.
This research uses quantitative research methodology.The research subjects were 298 students of the Information Systems Study Program at Indraprastha University PGRI Jakarta, Odd Semester, Academic Year 2023/2024, in the Basic Mathematics Course.The number of problem-solving skills questions is 15 questions.Data collection was carried out during the Odd Semester Midterm Examination (UTS) for the 2023/2024 Academic Year in October 2023.

Figure
Figure 1.Problem-Solving Skills Test Information Function (Bariyyah, 2021)orea, 2015)basic skills that need to be developed by training students regarding 21st-century learning(Nadia Mirela Florea, 2015).The results of research related to the problem-solving skills of Science Learning for Junior High School (SMP) students show that (I D Franestian, 2020): sub-skills identify problems by 5.2%, sub-skills cause and effect by 57.6%, sub-skills planning actions by 47.1%, determining relevant solutions by 37.6%, and 39.1% analyzing the influence of solutions.Based on this data, problem-solving skills are in the low category.Problem-solving skills consist of sub-skills: identifying problems, searching for and selecting variations of alternative solutions, and making problem-solving decisions(Bariyyah, 2021).Research results: 2.3% of problem-solving skills of the student's ability level were low, 62.3% of problem-solving skills of the student's ability level was moderate, 35.3% of problem- Prima: Jurnal Pendidikan Matematika ◼ 469 Analysis of Person'S Problem-Solving Skills Munali, Liberna, Mutakin solving skills of the student's ability level was high, and there are no students with very low levels of problem-solving skills.Based on this data, the level of problem-solving skills of the students in the medium category is 62.3%.Based on several theoretical studies related to problem-solving skills, this research is a measurement of the student's ability to respond to problem-solving skills questions with different dimensions or sub-skills.Problem-solving skills consist of two dimensions (Md, 2019): 1) observation skills (ability to observe) and 2) critical thinking skills (ability to think critically).

Table 1 . Interpretation of Cronbach's Alpha Score Reliability Cronbach's Alpha Score Reliability
(Sharifah Nurulhuda Tuan Mohd Yasin, 2018)ures, namely item reliability and person reliability(Sharifah Nurulhuda Tuan Mohd Yasin, 2018).The reliability of question items is the consistency of the assessment results produced by the instrument when tested on other groups of respondents with the same characteristics.Person reliability refers to how reliable a group of respondents is in providing the same response to another instrument that has items with the same level of difficulty.The reliability of Cronbach's Alpha score can be seen in Table1 below:

Table 4 . Person Reliability
Based on Table4, the Person Alpha Cronbach reliability coefficient score (KR-20) is 0.77.The person's holistic reliability coefficient score can be categorized as good and acceptable or the person's consistency in responding to the question items holistically is in the good and acceptable category.The person's reliability score is 0.85 means that the person's reliability can be categorized as good and acceptable.